Artificial intelligence for predicting adherence to CPAP therapy in obstructive sleep apnoea: A systematic review

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Sleep Medicine

Background:

• Continuous positive airway pressure (CPAP) therapy is crucial for managing obstructive sleep apnoea (OSA).
• Patient adherence to CPAP therapy is a significant clinical challenge impacting treatment effectiveness.
• Predicting adherence is essential for personalized OSA management strategies.

Purpose of the Study:

• To systematically review the application of artificial intelligence (AI) models for predicting CPAP adherence in adults with OSA.
• To evaluate the performance and methodologies of AI models used for this prediction.
• To identify gaps and future directions in AI-driven CPAP adherence prediction.

Main Methods:

• Systematic literature review adhering to PRISMA guidelines.
• Searched PubMed (MEDLINE) and IEEE Xplore for studies from January 2015 to March 2026.
• Included original studies and conference proceedings utilizing AI for CPAP adherence prediction in OSA patients, assessing risk of bias with PROBAST.

Main Results:

• Included 5 studies out of 104 identified records, noting heterogeneity in sample size and methodology.
• AI models (e.g., logistic regression, SVM, random forest, neural networks) showed moderate to good predictive performance (AUC up to 0.84, F1-score up to 0.86).
• All studies had risk-of-bias concerns, and none performed external validation.

Conclusions:

• AI demonstrates potential for predicting CPAP adherence and enabling personalized OSA management.
• The current evidence base is limited and heterogeneous, requiring cautious interpretation.
• Further research with external validation and real-world clinical evaluation is necessary for implementation.